When secondary replication is being performed, as shown in the following figure, both databases are up and running and all secondary transactions are being replicated to the primary database.

Figure 1. Secondary replication continues (before transition failover)


At this point, the secondary is considered the production database and the primary is considered the replica.

To perform the failover transition, schedule a time when no users are connected to the source or target. Users connected at the time of failover transition will be disconnected automatically, and their uncommitted transactions will be rolled back at the start of the transition process. No users may log in to the databases during the failover transition. In cases where the online-transition is set to 1, privileged read-only utilities may connect to the database during failover transition for reporting purposes, such as the dsrutil -C monitor utility.

The following figure illustrates the scheduling of database downtime.

Figure 2. Scheduling downtime to perform failback

To initiate the failback process using transition failover, issue the following command on the second machine:

dsrutil secondary-db-name -C transition failover

This command instructs the secondary database to begin transition. The failover command modifier instructs OpenEdge Replication that this is a failover transition and causes Replication to transition both the primary and secondary databases.

When the transition process begins, the Replication server informs the Replication agent on the primary machine to begin preparing the primary database for transition. If the online-transition property is set to 1, the database brokers stay online during the transition. If online-transition is not available, or is explicitly set to 0, the database brokers will perform multiple shutdowns and restarts during the failover transition.

Once the startup synchronization process is complete, the actual transition process can begin, as shown in the following figure.

Figure 3. Initiating transition of the primary database

Once the startup synchronization process completes normally, the transition of the primary database is performed as configured, as shown in the following figure.

Figure 4. Continuing transition of the primary database

Once the transition of the primary database reaches a critical point (that is, immediately before the database is to be restarted), the transition of the secondary database is performed. Once the transition of the secondary database completes normally, the completion of the transition of the primary database is started.

Once transition completes normally for both databases, the replication agent and servers terminate and restart, as shown in the following figure.

Figure 5. Databases started in new roles

As shown in the following figure, the roles of both databases have been reversed. The primary database is again the production database and the secondary database is again the replica. Primary replication is again being performed as it was before the initial failure occurred. The Replication server is replicating all primary transactions to the secondary database.

Figure 6. Primary replication activity is occurring